High Leukemic Stem Cell Frequency in FLT3-Mutated Acute Myeloid Leukemia Patients with Normal Karyotype: Implications for Prognosis

Normal karyotype in Acute Myeloid Leukemia (NK-AML) occurs in 40-50% of patients. Leukemic Stem Cells (LSC) persistence is known to trigger relapse, but the prognostic impact of their identification is understudied. We correlated LSC frequency at diagnosis with mutational profile and clinical outcomes in a Brazilian NK-AML cohort. Bone marrow samples (n=93) from 7 centers were submitted to LSC (defined as CD34+/CD38-/low/CD123+) quantification by multiparametric flow cytometry (MPFC). LSC quantification was correlated with clinical-laboratorial diagnosis data, ELN2017 risk stratification, FLT3 and NPM1 mutations and survival outcomes. Patients were categorized based on LSC percentage as lower (LSClow) or higher (LSChigh) than 0.1%. No difference in blast percentages or WBC counts between the LSC groups was found. Regarding risk stratification, the LSClow group was composed by (n=19, 59.4%) favorable, (n=13, 40.6%) intermediate and no adverse-risk individuals. In the LSChigh group, the patients were classified as favorable (n=25, 40.98%), intermediate (n=28, 45.9%), and adverse (n=6, 9.83%), thus suggesting that lower LSC frequency is associated with favorable risk, although statistical significance was not reached. In the LSClow group, 12.5% had the FLT3 mutation whereas in 33.3% of the LSChigh samples FLT3 mutation was detected. Of note, LSChigh samples presented lower FLT3 allelic ratio as compared to LSClow (0.47±0.28 vs 0.87±0.53). When associating LSC frequency with NPM1 gene status, 65.62% of LSClow group and 40% of LSChigh group were NPM1 mutated. When the co-mutational status was considered, among individuals with LSClow, 81.8% (N=18) was FLT3^wtNPM1^mut, 13.64% (N=3) was FLT3^mutNPM1^mut, and 4.56% (N=1) FLT3^mutNPM1^wt. In the LSChigh group, 37.5% (N=12) was FLT3^wtNPM1^mut, 37.5% (N=12) was FLT3^mutNPM1^mut, and 25% (N=8) FLT3^mutNPM1^mut. Overall, complete remission (CR/CRi) rate was 64.5% (N=60) after 2 cycles of chemotherapy. No significant difference was observed between groups: 59.4% (N=19) in the LSClow and 67.2% (N=41) in the LSChigh group achieved CR. After remission, 15.6% (N=5) of the patients relapsed in the LSClow group as compared to 16.2% (N=16) of the LSChigh patients. The LSClow group had a mean overall survival (OS) and event-free survival (EFS) of 568 and 541 days, respectively, whereas the LSChigh group had both mean OS and EFS of 466 days. We also assessed outcomes according to ELN2017 risk stratification. Favorable-risk patients with low LSC levels had a mean OS of 130 days, compared to 182 days for those with high LSC levels. Intermediate-risk patients with low LSC levels had a mean OS of 174 days, while the high LSC group had a mean OS of 228 days. Adverse-risk patients with low LSC levels had no measurable mean OS, whereas those with high LSC levels had a mean OS of 216 days. The Log-Rank test showed no significant differences in survival distributions between the groups. In summary, our data indicated that FLT3 mutation was more prevalent in patients with higher frequency of LSCs, indicating that cases with high LSC content at the flow cytometry initial examination should be quickly assessed for FLT3 mutation, especially considering that target therapy with FLT3 inhibitors can be included in the induction regimen. In agreement, ELN2017 favorable-risk patients had lower LSC quantification at diagnosis. The implication of LSC quantification in survival was not statistically demonstrated due to technical limitations including the number of patients with available data, but remission and survival rates suggested that LSChigh group present worse outcomes. In addition, LSC quantification by MPFC was feasible and can be easily incorporated to diagnosis and measurable residual disease panels. Indeed, the addition of more specific markers to define LSC could improve clinical stratification. The findings suggest that LSC identification is a potential biomarker that can guide clinicians on prognosis, particularly for assessment of AML without karyotypic abnormalities that rely only on mutational profiles, not always available in low and middle-income countries.

Pagnano:EMS: Research Funding; Pfizer: Speakers Bureau; Novartis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Teva: Speakers Bureau; Pintpharma: Speakers Bureau.